Visualization of Data for Prediction of Features for Human Eye Complications in Machine Learning Framework

In today’s world, where things have started progressing toward online medium, one of the major age groups exposed to this transition is that of high school and university students. This change has been accelerated especially during the pandemic, with the commencement of online classes, resulting in increased exposure of students to digital screens. In this paper, a comprehensive study as well as research has been conducted regarding the factors which contribute to the most common ocular problem among students that is myopia. The research begins with collection of data regarding input parameters like the number of hours spent in front of screens, level of dryness experienced in the eyes, age and gender through a questionnaire circulated in various educational institutions as the part of a survey. The highest number of responses is made up of students who are between the ages of 18 and 20 (83 responses), followed by those who are above 20 (74 responses), 15–18 (42 responses), and 12–15 (1 response). Further steps involve cleaning, analyzing, and visualizing the gathered data to derive hidden patterns and trends among various responses provided by the respondents. Nearly half of the pupils have myopia, and the majority (57.5%) spend excessive amounts of time using VDTs—more than four hours. Both genders have almost equal odds of developing near-sightedness or experiencing it already. 96% of the students report not regularly having dry eyes. In terms of dry eye, more than half of the students in levels 2, 3, and 4 are myopic. This indicates that the two variables have a somewhat positive relationship. It is a crucial step that builds a solid foundation for determining the form of input suitable for the successive stage of the process, which is the machine learning framework. This has been implemented using ML algorithms like Logistic regression, K nearest Neighbor, Decision tree, Random Forest, and Support Vector Machine.